2

I see a lot of 408 errors in my apache2 logs. I increased RequestReadTimeout and KeepAliveTimout but errors are still there.

The errors look like this:

xx.xx.xx.xx - - [05/Dec/2012:19:33:56 +0100] "-" 408 4561 "-" "-"
xx.xx.xx.xx - - [05/Dec/2012:19:33:56 +0100] "-" 408 4561 "-" "-"

I heard that these errors are related to Chrome optimization and some users did reported our site returning 408 internal error.

It is interesting that we get two 408 error from same IP in sequence then it that IP start working.

Improve this question

1 Answer 1

Reset to default
2

Those tend to be from web browsers doing predictive optimization; if the browser thinks there's a good chance the user is going to do something, it will prepare to do it. If you're the first result in someone's Google search, or their mouse cursor hovers over a link to your site a little too long, it may open a connection or start loading the page in anticipation of a click. http://forum.linode.com/viewtopic.php?p=47596

I would also read this answer on server fault which suggests not to mess with those apache settings https://serverfault.com/a/383294/94412

Improve this answer
2
  • The problem with explanation that this is "predictive optimization" and some customers do get this error.
    – Robert Grezan
    Dec 5, 2012 at 19:24
  • 1
    I'd run wireshark, or tail -f your apache logs and in Chrome or FireFox begin typing your domain name and see what happens in the logs once you select your domain name but don't load it in the browser.
    – Anagio
    Dec 5, 2012 at 19:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.